Answer: stop-sequences

## Explanation **stop-sequences** is the correct parameter to use when you want to stop text generation at specific phrases or sequences. ### Why stop-sequences is correct: 1. **Purpose**: Stop sequences allow you to define specific strings or phrases where the model should stop generating text 2. **Use case**: When you want to end generation at a particular phrase like "Thank you for contacting us." 3. **Functionality**: The model will stop generating once it encounters any of the specified stop sequences ### Why other options are incorrect: - **B. temperature**: Controls randomness in generation (higher = more random, lower = more deterministic) - **C. top-p**: Controls diversity via nucleus sampling (probability mass threshold) - **D. max-tokens**: Limits the maximum number of tokens generated, but doesn't stop at specific phrases ### Practical Application: In banking chatbots, using stop sequences ensures that responses end cleanly with standard closing phrases, preventing awkward or unnecessary continuation of conversation after the intended endpoint.

Author: Ritesh Yadav